---
title: "Presyo ng DeepSeek V4.1 Flash API: paghahambing ng 23 na provider"
description: "Presyo ng DeepSeek V4.1 Flash API kada token: pinakamababang nakalistang US$0.07 input at US$0.42 output kada 1M token."
url: https://broadrouter.com/fil/llm-api-pricing/deepseek-v4.1-flash/
lang: fil
updated: 2026-11-16
publisher: BroadRouter (a BroadNet site); API by BroadNet Inference, https://inference.broadnet.ai
---

# Presyo ng DeepSeek V4.1 Flash API

Na-update 2026-11-16

Sa 23 na provider na sinuri namin (binasa Nobyembre 14, 2026 hanggang Nobyembre 16, 2026), ang pinakamababang blended na standard na presyo para sa DeepSeek V4.1 Flash ay sa Relace: US$0.07 input at US$0.50 output kada 1M token. Naniningil ang sariling API ng DeepSeek ng US$0.30 / US$1.20 sa peak hours para sa bersyong sine-serve nito.

Salin ang page na ito ng [orihinal na Ingles](https://broadrouter.com/llm-api-pricing/deepseek-v4.1-flash/). Ang mga presyo, pangalan ng provider, petsa at quote mula sa source ay parehong data ng page sa Ingles at sabay na ina-update; nananatiling Ingles ang mga quote at tala ng provider.

DeepSeek · slug `deepseek-v4.1-flash` · hanggang 1,048,576 token ng context · 23 na provider na may standard na presyo. **Binasa ang bawat presyo mula sa sariling pricing page ng provider** at naka-link sa row nito, sinuri noong Nobyembre 16, 2026. Walang aggregator feed. [I-download ang lahat ng presyo (Ingles)](https://broadrouter.com/data/llm-api-provider-prices/) (bawat provider at model, may source, CSV o JSON). Kumpara sa buong market: tingnan ang [LLM API price index (Ingles)](https://broadrouter.com/data/llm-api-price-index/). Sariling site ng BroadNet ang BroadRouter.

## Magkano ang singil ng DeepSeek para sa DeepSeek V4.1 Flash?

Input

US$0.30

Cached input

US$0.006

Output

US$1.20

Blended 3:1

US$0.525

USD kada 1M token sa DeepSeek, model `deepseek-flash`, standard tier, peak hours; off-peak: US$0.15 input, US$0.003 cached at US$0.60 output. **Bersyong sine-serve:** Sine-serve ng API ng DeepSeek ang [DeepSeek V4.1 Flash 1017](https://broadrouter.com/fil/llm-api-pricing/deepseek-v4.1-flash-1017/) sa ilalim ng pangalang ito, kaya ipinapakita rito ang presyo nito bilang sanggunian at hindi ito nira-rank laban sa mga host ng DeepSeek V4.1 Flash. [Source](https://api-docs.deepseek.com/quick_start/pricing), sinuri noong Nobyembre 14, 2026.

**Mas mura ba ang isang host kaysa sa DeepSeek?** Oo. Naniningil ang Relace ng US$0.07 input at US$0.50 output kada 1M token, 66% na mas mababa kaysa sa sariling presyo ng DeepSeek na US$0.30 / US$1.20 sa peak hours sa 3:1 blend. Ang row ng DeepSeek ay para sa bersyong sine-serve ng API nito sa ilalim ng pangalang ito (ang 1017 snapshot ng DeepSeek V4.1 Flash), kaya puwede itong bahagyang maiba sa sine-serve ng mga host. Sa off-peak, naniningil ang DeepSeek ng US$0.15 input, US$0.003 cached at US$0.60 output.

## Mga detalye ng model: DeepSeek V4.1 Flash

Mga detalye mula sa model card ng DeepSeek para sa [`deepseek-ai/DeepSeek-V4.1-Flash`](https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash) at sa Hugging Face model API, binasa noong Nobyembre 16, 2026. Hindi ito bineripika ng BroadNet, at puwedeng i-serve ng mga host ang model sa ibang precision. Site ng BroadNet ang BroadRouter.

DeepSeek V4.1 Flash: mga detalye ayon sa model card at sa Hugging Face model API, may mga source
| Detalye | Gaya ng nakasaad | Source |
| --- | --- | --- |
| Open weights | Oo. Nasa pampublikong Hugging Face repository (`deepseek-ai/DeepSeek-V4.1-Flash`) ang mga weight, nang walang access gate. | [Hugging Face model API](https://huggingface.co/api/models/deepseek-ai/DeepSeek-V4.1-Flash), sinuri noong Nobyembre 16, 2026 · Quote: "private": false, "gated": false, 48 .safetensors files listed |
| Lisensya | MIT (mit) | [metadata ng model card ng DeepSeek](https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash), sinuri noong Nobyembre 16, 2026 · Quote: license: mit |
| Kabuuang parameter | 552B backbone parameters (the card lists 196B of Engram conditional memory separately) | [model card ng DeepSeek](https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash), sinuri noong Nobyembre 16, 2026 · Quote: a multimodal Mixture-of-Experts (MoE) model with 552B backbone parameters … Engram conditional memory (196B parameters, sparsely accessed via token-based lookup) |
| Aktibong parameter kada token | 8B per token in prefill, 16B in decode | [model card ng DeepSeek](https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash), sinuri noong Nobyembre 16, 2026 · Quote: activate only 8B parameters per token during prefill and 16B during decode |
| Haba ng context | Up to one million tokens | [model card ng DeepSeek](https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash), sinuri noong Nobyembre 16, 2026 · Quote: support for contexts of up to one million tokens |
| Input at output | Images and text in; text out | [model card ng DeepSeek](https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash), sinuri noong Nobyembre 16, 2026 · Quote: The model natively processes images and text, and generates text autoregressively. |
| Tool calling | Yes, per the prompt-encoding reference | [model card ng DeepSeek](https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash), sinuri noong Nobyembre 16, 2026 · Quote: test cases for multi-turn conversations, tool calling, thinking mode |
| Unang nailathala sa Hugging Face (hindi kinakailangang petsa ng release) | Setyembre 10, 2026 | [Hugging Face model API](https://huggingface.co/api/models/deepseek-ai/DeepSeek-V4.1-Flash), sinuri noong Nobyembre 16, 2026 · Quote: "createdAt": "2026-09-10T02:17:58.000Z" |

## Presyo ng DeepSeek V4.1 Flash ayon sa provider

USD kada 1M token. Nira-rank ang mga standard na presyo sa blend ng 3 input token sa 1 output token, pinakamura muna; ang pinakamahal na standard na presyo ay 3.0× ng pinakamura. Kasunod ang mga libreng row, ang bersyong sine-serve ng gumawa at ang iba pang service tier, may tag at hindi naka-rank.

Presyo ng DeepSeek V4.1 Flash API ayon sa provider
| # | Provider | Input | Cached input | Output | Blended 3:1 | Context | Precision | Source |
| --- | --- | --- | --- | --- | --- | --- | --- | --- |
| 1 | **Relace** pinakamura · Quote: DeepSeek V4.1 Flash $0.07 $0.008 $0.50 | US$0.07 | US$0.008 | US$0.50 | US$0.1775 | hindi nakasaad | hindi nakasaad | [page ng Relace](https://relace.ai/pricing) · sinuri noong Nobyembre 16, 2026 |
| 2 | **River** · `deepseek-ai/DeepSeek-V4.1-Flash` · Note: Preview rate. Quote: DeepSeek-V4.1-Flash $0.12 / 1M $0.0024 / 1M $0.48 / 1M | US$0.12 | US$0.0024 | US$0.48 | US$0.21 | hindi nakasaad | hindi nakasaad | [page ng River](https://river.ai/pricing) · sinuri noong Nobyembre 16, 2026 |
| 3 | **DeepInfra** promo · `deepseek-ai/DeepSeek-V4.1-Flash` · Note: Promotional price, 30% off list (list 0.2 in / 0.6 out). Quote: $0.14 in $0.42 out $0.004 cached / 1M tokens | US$0.14 | US$0.0042 | US$0.42 | US$0.21 | 1,048,576 | FP8 | [page ng DeepInfra](https://deepinfra.com/deepseek-ai/DeepSeek-V4.1-Flash) · sinuri noong Nobyembre 14, 2026 · muling sinuri noong Nobyembre 16, 2026 |
| 4 | **CoreWeave Forge** · `deepseek-ai/DeepSeek-V4.1-Flash` · Quote: DeepSeek V4.1-Flash $0.13 $0.65 $0.002 | US$0.13 | US$0.002 | US$0.65 | US$0.26 | 1,048,576 | hindi nakasaad | [page ng CoreWeave Forge](https://coreweave.com/forge-pricing/inference-agents) · sinuri noong Nobyembre 15, 2026 · muling sinuri noong Nobyembre 16, 2026 |
| 5 | **Doubleword** · Note: Realtime priority. Quote: DeepSeek V4.1 Flash $0.15 / $0.01 / $0.60 $0.12 / $0.01 / $0.48 $0.08 / $0.01 / $0.30 | US$0.15 | US$0.01 | US$0.60 | US$0.2625 | hindi nakasaad | hindi nakasaad | [page ng Doubleword](https://docs.doubleword.ai/inference-api/models) · sinuri noong Nobyembre 16, 2026 |
| 6 | **SiliconFlow** · Note: Context shown as 1049K. Quote: DeepSeek-V4.1-Flash 1049K $0.15 $0.003 $0.6 | US$0.15 | US$0.003 | US$0.60 | US$0.2625 | 1,048,576 | hindi nakasaad | [page ng SiliconFlow](https://www.siliconflow.com/pricing) · sinuri noong Nobyembre 14, 2026 · muling sinuri noong Nobyembre 16, 2026 |
| 7 | **Morph** · `morph-dsv41flash` · Quote: DeepSeek V4.1 Flash morph-dsv41flash $ 0.12 /1M $ 0.008 /1M $ 0.80 /1M 1M | US$0.12 | US$0.008 | US$0.80 | US$0.29 | 1,048,576 | hindi nakasaad | [page ng Morph](https://www.morphllm.com/pricing) · sinuri noong Nobyembre 16, 2026 |
| 8 | **GMI Cloud** promo · `deepseek-ai/DeepSeek-V4.1-Flash` · Note: Promotional price, 40% off list (list 0.3 in / 1.2 out / 0.006 cached). Quote: Price (per 1M tokens) 40% off Input $0.18 $0.30 Output $0.72 $1.20 Cache Read $0.0036 $0.006 Cache Write - | US$0.18 | US$0.0036 | US$0.72 | US$0.315 | 1,048,575 | FP8 | [page ng GMI Cloud](https://console.gmicloud.ai/model/llm/deepseek-v4-1-flash/0049392b-5a6b-4e94-acc0-e18b8e293aac) · sinuri noong Nobyembre 15, 2026 · muling sinuri noong Nobyembre 16, 2026 |
| 9 | **Inference.net** · `deepseek-v4.1-flash` · Quote: DeepSeek V4.1 Flash 1M input context Vision Tools Reasoning $0.30 $0.007 $1.00 deepseek-v4.1-flash | US$0.30 | US$0.007 | US$1 | US$0.475 | 1,040,000 | FP8 | [page ng Inference.net](https://inference.net/models) · sinuri noong Nobyembre 16, 2026 |
| 10 | **Alibaba Cloud** · `deepseek-v4.1-flash` · Note: International (Singapore) deployment. Peak (busy-hour) price; busy hours are 08:00–22:00 UTC+8 (00:00–14:00 UTC), idle hours bill 0.15 input, 0.6 output. Cached price from the Context Cache page: "a fixed $0.003 per 1M tokens during off-peak hours and $0.006 per 1M tokens during peak hours". Quote: deepseek-v4.1-flash context caching discount International Busy hours: $0.3 Idle hours: $0.15 Busy hours: $1.2 Idle hours: $0.6 | US$0.30 | US$0.006 | US$1.20 | US$0.525 | hindi nakasaad | hindi nakasaad | [page ng Alibaba Cloud](https://www.alibabacloud.com/help/en/model-studio/model-pricing) · sinuri noong Nobyembre 14, 2026 · muling sinuri noong Nobyembre 16, 2026 |
| 11 | **Baseten** · `deepseek-ai/DeepSeek-V4.1-Flash` · Note: Context from the model library page (https://www.baseten.co/library/deepseek-v41-flash/), stated as 1M. Quote: DeepSeek V4.1 Flash $0.30 $0.007 $1.20 | US$0.30 | US$0.007 | US$1.20 | US$0.525 | 1,000,000 | hindi nakasaad | [page ng Baseten](https://www.baseten.co/pricing/) · sinuri noong Nobyembre 14, 2026 · muling sinuri noong Nobyembre 16, 2026 |
| 12 | **DigitalOcean** · Note: DigitalOcean-hosted. Quote: DeepSeek V4.1 Flash — Processing modePrompt lengthInputOutputCache read: Standard All prompts$0.30$1.20$0.006 | US$0.30 | US$0.006 | US$1.20 | US$0.525 | hindi nakasaad | hindi nakasaad | [page ng DigitalOcean](https://docs.digitalocean.com/products/inference/details/pricing/) · sinuri noong Nobyembre 14, 2026 · muling sinuri noong Nobyembre 16, 2026 |
| 13 | **Featherless** · `deepseek-ai/DeepSeek-V4.1-Flash` · Note: Developer plan (per-token billing, from 50 USD a month in credits). Quote: PRICING: INPUT $0.3 / CACHED $0.03 / OUTPUT $1.2 | US$0.30 | US$0.03 | US$1.20 | US$0.525 | 262,144 | FP8 | [page ng Featherless](https://featherless.ai/models/deepseek-ai/DeepSeek-V4.1-Flash) · sinuri noong Nobyembre 15, 2026 · muling sinuri noong Nobyembre 16, 2026 |
| 14 | **Fireworks AI** · `accounts/fireworks/models/deepseek-v4p1-flash` · Quote: DeepSeek V4.1 Flash \| $0.30 / $0.006 / $1.20 \| $0.375 / $0.0075 / $1.50 | US$0.30 | US$0.006 | US$1.20 | US$0.525 | 1,048,576 | hindi nakasaad | [page ng Fireworks AI](https://docs.fireworks.ai/serverless/pricing) · sinuri noong Nobyembre 14, 2026 · muling sinuri noong Nobyembre 16, 2026 |
| 15 | **Google Vertex AI** · `deepseek-v4.1-flash` · Note: Google says it was free until October 16, 2026; this is the price after that date. Quote: DeepSeek-V4.1-Flash * Input $0.30 … Output $1.20 … Cache Hit $0.03 | US$0.30 | US$0.03 | US$1.20 | US$0.525 | hindi nakasaad | hindi nakasaad | [page ng Google Vertex AI](https://cloud.google.com/model-garden/partner-models/pricing) · sinuri noong Nobyembre 14, 2026 · muling sinuri noong Nobyembre 16, 2026 |
| 16 | **Hyperbolic** · `deepseek-v4.1-flash` · Note: Context shown on page as "1M" (rounded); stored as the decimal value. Quote: DeepSeek V4.1 Flash deepseek-v4.1-flash DeepSeek · 1M context · Text $0.30 $1.20 $0.006 | US$0.30 | US$0.006 | US$1.20 | US$0.525 | 1,000,000 | hindi nakasaad | [page ng Hyperbolic](https://www.hyperbolic.ai/inference) · sinuri noong Nobyembre 14, 2026 · muling sinuri noong Nobyembre 16, 2026 |
| 17 | **Microsoft Azure** · Note: Foundry Models serverless (pay-as-you-go), Global deployment. Also sold as "FW DeepSeek V4.1 Flash" (Fireworks on Azure, Global): input $0.375, cached $0.008, output $1.50. Quote: DeepSeek V4.1 Flash Global $0.30 $0.006 $1.20 | US$0.30 | US$0.006 | US$1.20 | US$0.525 | hindi nakasaad | hindi nakasaad | [page ng Microsoft Azure](https://azure.microsoft.com/en-us/pricing/details/ai-foundry-models/deepseek/) · sinuri noong Nobyembre 14, 2026 · muling sinuri noong Nobyembre 16, 2026 |
| 18 | **Modal** · Quote: DeepSeek V4.1 Flash Prompt: $0.30 Cached prompt: $0.03 Completion: $1.20 per 1M tokens. | US$0.30 | US$0.03 | US$1.20 | US$0.525 | hindi nakasaad | hindi nakasaad | [page ng Modal](https://modal.com/pricing) · sinuri noong Nobyembre 16, 2026 |
| 19 | **Novita AI** · `deepseek/deepseek-v4.1-flash` · Quote: DeepSeek V4.1 Flash 1M $0.3 /Mt · Cache Read $0.006 /Mt $1.2 /Mt | US$0.30 | US$0.006 | US$1.20 | US$0.525 | 1,048,576 | hindi nakasaad | [page ng Novita AI](https://novita.ai/pricing) · sinuri noong Nobyembre 14, 2026 · muling sinuri noong Nobyembre 16, 2026 |
| 20 | **Parasail** · Quote: DeepSeek V4.1 Flash $0.30 $1.20 $0.006 | US$0.30 | US$0.006 | US$1.20 | US$0.525 | hindi nakasaad | hindi nakasaad | [page ng Parasail](https://www.parasail.io/pricing) · sinuri noong Nobyembre 14, 2026 · muling sinuri noong Nobyembre 16, 2026 |
| 21 | **Reka** · `deepseek41-flash` · Quote: deepseek41-flash $0.30/M input tokens $1.20/M output tokens $0.006/M cached | US$0.30 | US$0.006 | US$1.20 | US$0.525 | 1,048,576 | hindi nakasaad | [page ng Reka](https://developer.reka.ai/models) · sinuri noong Nobyembre 16, 2026 |
| 22 | **Together AI** · Note: US geo-pinned option: $0.375 in / $1.50 out. Quote: DeepSeek V4.1 Flash $0.30 $0.006 (cached) $1.20 | US$0.30 | US$0.006 | US$1.20 | US$0.525 | hindi nakasaad | hindi nakasaad | [page ng Together AI](https://www.together.ai/pricing) · sinuri noong Nobyembre 14, 2026 · muling sinuri noong Nobyembre 16, 2026 |
| 23 | **Venice** · `deepseek-v4-1-flash` · Quote: \| DeepSeek V4.1 Flash \| `deepseek-v4-1-flash` \| $0.30 \| $1.20 \| $0.0075 \| — \| Private \| | US$0.30 | US$0.0075 | US$1.20 | US$0.525 | 1,000,000 | FP8 | [page ng Venice](https://docs.venice.ai/pricing) · sinuri noong Nobyembre 15, 2026 · muling sinuri noong Nobyembre 16, 2026 |
| — | **DeepSeek** gumawa, bersyong sine-serve presyo sa peak · `deepseek-flash` · Note: Peak-hour price (first price column of the quoted table). Off-peak is half: 0.15 input, 0.003 cached, 0.6 output. Peak hours are 01:00 - 04:00 and 06:00 - 10:00 UTC, Monday through Friday, excluding Chinese public holidays. All other hours are off-peak. API model name deepseek-flash serves DeepSeek-V4.1-Flash-1017; the legacy names deepseek-v4-flash and deepseek-v4-flash-vision-exp are retired and served by this model at the Flash price. Context 1M, max output 384K. Quote: (CACHE HIT) OFF-PEAK $0.003 $0.022 PEAK $0.006 $0.044 1M INPUT TOKENS (CACHE MISS) OFF-PEAK $0.15 $0.66 PEAK $0.3 $1.32 1M OUTPUT TOKENS OFF-PEAK $0.6 $1.98 PEAK $1.2 $3.96 | US$0.30 | US$0.006 | US$1.20 | US$0.525 | 1,000,000 | hindi nakasaad | [page ng DeepSeek](https://api-docs.deepseek.com/quick_start/pricing) · sinuri noong Nobyembre 14, 2026 · muling sinuri noong Nobyembre 16, 2026 |
| — | **Doubleword** batch · Note: Batch (24h) priority. Quote: DeepSeek V4.1 Flash $0.15 / $0.01 / $0.60 $0.12 / $0.01 / $0.48 $0.08 / $0.01 / $0.30 | US$0.08 | US$0.01 | US$0.30 | US$0.135 | hindi nakasaad | hindi nakasaad | [page ng Doubleword](https://docs.doubleword.ai/inference-api/models) · sinuri noong Nobyembre 16, 2026 |
| — | **DeepInfra** flex / spot promo · `deepseek-ai/DeepSeek-V4.1-Flash` · Note: Flex service tier (0.8× standard). Promotional price, 30% off list (list 0.2 in / 0.6 out). Quote: Flex (0.8×) $0.112 $0.336 $0.0034 | US$0.112 | US$0.0034 | US$0.336 | US$0.168 | 1,048,576 | FP8 | [page ng DeepInfra](https://deepinfra.com/deepseek-ai/DeepSeek-V4.1-Flash) · sinuri noong Nobyembre 14, 2026 · muling sinuri noong Nobyembre 16, 2026 |
| — | **Doubleword** flex / spot · Note: Async priority. Quote: DeepSeek V4.1 Flash $0.15 / $0.01 / $0.60 $0.12 / $0.01 / $0.48 $0.08 / $0.01 / $0.30 | US$0.12 | US$0.01 | US$0.48 | US$0.21 | hindi nakasaad | hindi nakasaad | [page ng Doubleword](https://docs.doubleword.ai/inference-api/models) · sinuri noong Nobyembre 16, 2026 |
| — | **Fireworks AI** flex / spot · `accounts/fireworks/models/deepseek-v4p1-flash` · Note: Fireworks "Spot" mode: 30% off Standard on best-effort capacity; throughput can drop to zero. Quote: DeepSeek V4.1 Flash Spot $0.0042 $0.21 $0.84 | US$0.21 | US$0.0042 | US$0.84 | US$0.3675 | 1,048,576 | hindi nakasaad | [page ng Fireworks AI](https://fireworks.ai/pricing) · sinuri noong Nobyembre 14, 2026 · muling sinuri noong Nobyembre 16, 2026 |
| — | **DeepInfra** priority promo · `deepseek-ai/DeepSeek-V4.1-Flash` · Note: Priority service tier (1.5× standard). Promotional price, 30% off list (list 0.2 in / 0.6 out). Quote: Priority (1.5×) $0.21 $0.63 $0.0063 | US$0.21 | US$0.0063 | US$0.63 | US$0.315 | 1,048,576 | FP8 | [page ng DeepInfra](https://deepinfra.com/deepseek-ai/DeepSeek-V4.1-Flash) · sinuri noong Nobyembre 14, 2026 · muling sinuri noong Nobyembre 16, 2026 |
| — | **Fireworks AI** priority · `accounts/fireworks/models/deepseek-v4p1-flash` · Note: Priority mode. Quote: DeepSeek V4.1 Flash \| Priority $0.375 / $0.0075 / $1.50 | US$0.375 | US$0.0075 | US$1.50 | US$0.6563 | 1,048,576 | hindi nakasaad | [page ng Fireworks AI](https://docs.fireworks.ai/serverless/pricing) · sinuri noong Nobyembre 14, 2026 · muling sinuri noong Nobyembre 16, 2026 |
| — | **Baseten** fast · Note: Context from the model library page (https://www.baseten.co/library/deepseek-v41-flash-fast/), stated as 1M. Quote: DeepSeek V4.1 Flash Fast $0.60 $0.014 $2.40 | US$0.60 | US$0.014 | US$2.40 | US$1.05 | 1,000,000 | hindi nakasaad | [page ng Baseten](https://www.baseten.co/pricing/) · sinuri noong Nobyembre 14, 2026 · muling sinuri noong Nobyembre 16, 2026 |
| — | **BroadNet Inference** may-ari ng site na ito · `broadnet/deepseek-v4.1-flash` · Parehong model. Hindi naka-rank: sariling site ng BroadNet ang BroadRouter. [Page ng model](https://broadrouter.com/fil/models/deepseek-v4.1-flash/) | US$0.175 | US$0.0055 | US$0.52 | US$0.2612 | 1,048,576 | hindi nakasaad | [pricing.json (Ingles)](https://inference.broadnet.ai/pricing.json?utm_source=broadrouter&utm_medium=referral&utm_campaign=fil-llm-api-pricing-deepseek-v4.1-flash&utm_content=price-table) · na-sync noong Nobyembre 14, 2026 |

-   **Naka-rank:** standard-tier na presyo lang, ayon sa (3 × input + output) ÷ 4. Ang "Cached input: hindi nakalathala" ay nangangahulugang walang cache price sa page ng provider; i-budget ang mga cached token sa input price.
-   **Promo:** inilalarawan ng page ng provider bilang promotional o discounted ang presyo (DeepInfra at GMI Cloud). Nira-rank ito ayon sa pagkakalathala, pero puwede itong matapos nang walang abiso; nasa mga tala ng row ang list price o petsa ng pagtatapos kung nakasaad ito sa page.
-   **Iba pang tier (batch, flex / spot, priority, at fast):** tumatakbo nang asynchronous ang batch sa loob ng ilang oras; best-effort ang flex at spot at puwedeng bumagal o tanggihan kapag puno ang capacity; mas mahal ang priority at fast kapalit ng reliability o bilis. Ipinapakita, hindi nira-rank.
-   **Presyo sa peak:** may peak at off-peak rate ang DeepSeek; dala ng row ang presyo sa peak at nasa mga tala nito ang off-peak rate.
-   **Automated na muling pagsuri Nobyembre 16, 2026:** linggu-linggong kinukuha muli ng isang job ang bawat source page at hinahanap ang naitalang quote o presyo. Hindi ito review ng tao. Ang "muling sinuri" ay nangangahulugang nakita pa rin ang mga ito. Ang mga row na walang tala ng muling pagsuri ay nagpapakita lang ng petsa kung kailan binasa ang presyo.
-   **Precision:** ipinapakita lang kung nakasaad ito ng provider; ang "hindi nakasaad" ay hindi alam, hindi full precision.

[I-report ang maling presyo o numero](mailto:info@broadnet.ai?subject=Wrong%20price%20or%20number%20on%20DeepSeek%20V4.1%20Flash%20API%20pricing&body=Page%3A%20https%3A%2F%2Fbroadrouter.com%2Ffil%2Fllm-api-pricing%2Fdeepseek-v4.1-flash%2F%0A%0AWhat%20looks%20wrong%20(model%2C%20figure%20shown%2C%20correct%20figure%20and%20a%20source%20link%20if%20you%20have%20one)%3A%0A) (email sa info@broadnet.ai, nakalagay na ang link ng page na ito).

## Magkano ang DeepSeek V4.1 Flash para sa totoong workload?

Buwanang gastos sa loob ng 30 araw sa bawat standard na presyo, gamit ang mga preset ng [LLM API cost calculator (Ingles)](https://broadrouter.com/tools/llm-api-cost-calculator/). Sinisingil sa input price ang cached input ng mga provider na walang cached price.

-   **Support chatbot:** 2,000 request kada araw, 1,500 input token (60% cached), 300 output
-   **RAG sa mga dokumento:** 1,000 request kada araw, 8,000 input token (30% cached), 500 output
-   **Coding agent (isang developer):** 1,500 request kada araw, 40,000 input token (85% cached), 2,000 output kasama ang reasoning

Buwanang gastos ng tatlong workload sa DeepSeek V4.1 Flash, ayon sa provider
| Provider | Support chatbot | RAG sa mga dokumento | Coding agent (isang developer) |
| --- | --- | --- | --- |
| **Relace** | US$11.95 | US$19.84 | US$76.14 |
| **River** | US$13.09 | US$27.53 | US$79.27 |
| **DeepInfra** (promo) | US$12.83 | US$30.12 | US$82.03 |
| **CoreWeave Forge** | US$16.49 | US$31.73 | US$96.66 |
| **Doubleword** | US$16.74 | US$34.92 | US$109.80 |
| **SiliconFlow** | US$16.36 | US$34.42 | US$99.09 |
| **Morph** | US$19.15 | US$32.74 | US$116.64 |
| **GMI Cloud** (promo) | US$19.63 | US$41.30 | US$118.91 |
| **Inference.net** | US$29.18 | US$65.90 | US$181.71 |
| **Alibaba Cloud** | US$32.72 | US$68.83 | US$198.18 |
| **Baseten** | US$32.78 | US$68.90 | US$199.71 |
| **DigitalOcean** | US$32.72 | US$68.83 | US$198.18 |
| **Featherless** | US$34.02 | US$70.56 | US$234.90 |
| **Fireworks AI** | US$32.72 | US$68.83 | US$198.18 |
| **Google Vertex AI** | US$34.02 | US$70.56 | US$234.90 |
| **Hyperbolic** | US$32.72 | US$68.83 | US$198.18 |
| **Microsoft Azure** | US$32.72 | US$68.83 | US$198.18 |
| **Modal** | US$34.02 | US$70.56 | US$234.90 |
| **Novita AI** | US$32.72 | US$68.83 | US$198.18 |
| **Parasail** | US$32.72 | US$68.83 | US$198.18 |
| **Reka** | US$32.72 | US$68.83 | US$198.18 |
| **Together AI** | US$32.72 | US$68.83 | US$198.18 |
| **Venice** | US$32.81 | US$68.94 | US$200.47 |
| **DeepSeek** (gumawa, bersyong sine-serve, peak) | US$32.72 | US$68.83 | US$198.18 |
| **BroadNet Inference** (may-ari ng site na ito) | US$15.96 | US$37.60 | US$102.46 |

Para sa support chatbot, nakakatipid ang pinakamurang presyo (Relace, US$11.95 kada buwan) ng US$20.77 kumpara sa median (US$32.72). Para sa coding agent, US$122.04 kada buwan ang agwat (62% na mas mababa).

## DeepSeek V4.1 Flash sa BroadNet Inference

Oo. Sine-serve ng BroadNet Inference ang DeepSeek V4.1 Flash bilang `broadnet/deepseek-v4.1-flash` sa US$0.175 input, US$0.0055 cached at US$0.52 output kada 1M token (na-sync ang rate card noong Nobyembre 14, 2026), na may zero data retention. Sa 3:1 blend, 4 na standard na presyo sa page na ito ang mas mababa. Sariling site ng BroadNet ang BroadRouter. [DeepSeek V4.1 Flash sa BroadNet](https://broadrouter.com/fil/models/deepseek-v4.1-flash/).

Nasa [bersyong Ingles ng page na ito (Ingles)](https://broadrouter.com/llm-api-pricing/deepseek-v4.1-flash/) ang mga nailathalang benchmark score, gastos ayon sa wika at ang pagsusuri ng traffic mix.

## Tantiyahin ang sarili mong bill

-   [LLM API cost calculator na naka-preload ang DeepSeek V4.1 Flash (Ingles)](https://broadrouter.com/tools/llm-api-cost-calculator/?model=deepseek-v4.1-flash): ang traffic at cache rate mo, laban sa alinman sa 431 na model.

## Paano ikinukumpara ang DeepSeek V4.1 Flash sa ibang model ng DeepSeek?

Ang mga model ng DeepSeek na ang pinakamurang standard na presyo ay pinakamalapit sa DeepSeek V4.1 Flash (US$0.1775 blended):

Mga model ng DeepSeek na pinakamalapit ang presyo sa DeepSeek V4.1 Flash
| Model | Pinakamurang input | Pinakamurang output | Max context | Mga provider |
| --- | --- | --- | --- | --- |
| [DeepSeek R1 Distill Qwen 14B (Ingles)](https://broadrouter.com/llm-api-pricing/deepseek-r1-distill-qwen-14b/) | US$0.20 | US$0.20 | 131,072 | 2 |
| [DeepSeek R1 Distill Llama 8B (Ingles)](https://broadrouter.com/llm-api-pricing/deepseek-r1-distill-llama-8b/) | US$0.14 | US$0.39 | 131,072 | 2 |
| [DeepSeek V4.1 Flash 1017](https://broadrouter.com/fil/llm-api-pricing/deepseek-v4.1-flash-1017/) | US$0.09 | US$0.30 | 1,048,576 | 23 |
| [DeepSeek V3.2](https://broadrouter.com/fil/llm-api-pricing/deepseek-v3.2/) | US$0.2088 | US$0.3096 | 163,840 | 11 |
| [DeepSeek R1 Distill Qwen 7B (Ingles)](https://broadrouter.com/llm-api-pricing/deepseek-r1-distill-qwen-7b/) | US$0.10 | US$0.20 | 131,072 | 2 |
| [DeepSeek-V3.2-Exp (Ingles)](https://broadrouter.com/llm-api-pricing/deepseek-v3.2-exp/) | US$0.27 | US$0.41 | 163,840 | 2 |

Pinakamalapit ang presyo mula sa ibang gumawa: [Gemini 2.5 Flash-Lite (Ingles)](https://broadrouter.com/llm-api-pricing/gemini-2.5-flash-lite/) (Google, US$0.175) · [MiMo v2.5 (Ingles)](https://broadrouter.com/llm-api-pricing/mimo-v2.5/) (Xiaomi, US$0.175) · [Qwen3.8 Flash (Ingles)](https://broadrouter.com/llm-api-pricing/qwen3.8-flash/) (Qwen, US$0.1803).

[Lahat ng 431 na model: paghahambing ng presyo ng API](https://broadrouter.com/fil/llm-api-pricing/) · [Pinakamurang LLM API ngayon](https://broadrouter.com/fil/llm-api-pricing/cheapest/) · [LLM API price index (open data) (Ingles)](https://broadrouter.com/data/llm-api-price-index/) · [I-download ang lahat ng presyo (CSV, JSON) (Ingles)](https://broadrouter.com/data/llm-api-provider-prices/) · [Log ng pagbabago ng presyo (Ingles)](https://broadrouter.com/llm-api-pricing/changes/)

## Mga tanong

### Magkano ang DeepSeek V4.1 Flash API kada token?

Mula US$0.07 kada 1M input token (US$0.00000007 kada token) at US$0.42 kada 1M output token (US$0.00000042 kada token) sa pinakamababang standard na presyong nakita namin, na puwedeng magmula sa magkaibang provider. Ang median ng 23 na standard na presyo ay US$0.30 input at US$1.20 output. Binasa ang bawat presyo mula sa sariling pricing page ng provider; naka-link ang source sa row nito.

### Aling provider ang pinakamura para sa DeepSeek V4.1 Flash?

Sa 23 na provider na sinuri namin (binasa Nobyembre 14, 2026 hanggang Nobyembre 16, 2026), ang pinakamababa ay ang Relace, naka-rank sa blend ng 3 input token sa 1 output token (US$0.1775 kada 1M). Sa input lang, ang pinakamababang presyo ay US$0.07 (Relace); sa output lang, US$0.42 (DeepInfra). Ipinapakita sa table ang batch, flex, priority at fast tier pero hindi naka-rank.

### Mas mura ba ang DeepSeek V4.1 Flash sa isang host kaysa sa DeepSeek?

Oo. Naniningil ang Relace ng US$0.07 input at US$0.50 output kada 1M token, 66% na mas mababa kaysa sa sariling presyo ng DeepSeek na US$0.30 / US$1.20 sa peak hours sa 3:1 blend. Ang row ng DeepSeek ay para sa bersyong sine-serve ng API nito sa ilalim ng pangalang ito (ang 1017 snapshot ng DeepSeek V4.1 Flash), kaya puwede itong bahagyang maiba sa sine-serve ng mga host. Sa off-peak, naniningil ang DeepSeek ng US$0.15 input, US$0.003 cached at US$0.60 output.

### May cached input pricing ba ang DeepSeek V4.1 Flash?

May cached-input price ang lahat ng 23 na standard na presyo. Ang pinakamababa ay US$0.002 kada 1M sa CoreWeave Forge; ang median ay US$0.006. Sinisingil sa input price ang cached input ng mga provider na walang cached price. [Ano ang ibig sabihin ng cached input](https://broadrouter.com/fil/glossary/cached-input-tokens/).

### Gaano kalaki ang context window ng DeepSeek V4.1 Flash?

Hanggang 1,048,576 token. 6 na standard na presyo ang may mas maliit na context, pababa hanggang 262,144 token (Featherless, Baseten, Hyperbolic, Venice at 2 pa). Tingnan ang context column bago mo piliin ang pinakamurang row para sa mahahabang prompt. [Context window calculator (Ingles)](https://broadrouter.com/tools/context-window-calculator/).

### May batch, flex o priority na presyo ba para sa DeepSeek V4.1 Flash?

Oo: batch mula sa Doubleword sa US$0.08 / US$0.30; flex / spot mula sa DeepInfra sa US$0.112 / US$0.336; priority mula sa DeepInfra sa US$0.21 / US$0.63; fast mula sa Baseten sa US$0.60 / US$2.40 kada 1M input / output token. Ipinagpapalit ng mga tier na ito ang bilis o availability kapalit ng presyo (tumatakbo ang batch sa loob ng ilang oras, puwedeng bumagal o tanggihan ang flex at spot kapag puno ang capacity, mas mahal ang priority at fast kapalit ng bilis o reliability), kaya ipinapakita ang mga ito pero hindi nira-rank.

### Libre bang gamitin ang DeepSeek V4.1 Flash sa pamamagitan ng API?

Walang provider sa data namin na naglilista ng libreng presyo para sa DeepSeek V4.1 Flash, ayon sa pagsuri noong Nobyembre 16, 2026. Sine-serve ito ng BroadNet Inference bilang `broadnet/deepseek-v4.1-flash` sa mga may-bayad na account lang: hindi ito sakop ng US$5 na panimulang credit (na-sync ang rate card noong Nobyembre 14, 2026). Nakalista sa [Free tier at credit ayon sa provider (Ingles)](https://broadrouter.com/data/llm-api-free-tiers-and-rate-limits/) (sinuri noong Nobyembre 14, 2026) ang mga pangkalahatang alok na walang pinangalanang model; nasa iisang page ang [lahat ng libreng LLM API model](https://broadrouter.com/fil/llm-api-pricing/free/).

### Open source ba ang DeepSeek V4.1 Flash, may open weights ba ito, at ano ang lisensya nito?

Inilalathala ng DeepSeek ang mga weight sa Hugging Face bilang [`deepseek-ai/DeepSeek-V4.1-Flash`](https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash), sa pampublikong repository na walang access gate. Ang lisensya, ayon sa [metadata ng model card ng DeepSeek](https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash): MIT (mit). Ang ibig sabihin ng open weights ay puwede mong i-download at patakbuhin ang model nang ikaw mismo; ang lisensyang iyon ang nagtatakda kung ano ang puwede mong gawin dito, kasama ang commercial na paggamit, kaya basahin mo ito nang buo (sinuri noong Nobyembre 16, 2026). Mula sa card ng DeepSeek at sa Hugging Face model API ang mga detalyeng ito; hindi ito bineripika ng BroadNet, at puwedeng i-serve ng mga host ang model sa ibang precision.

### Ilan ang parameter ng DeepSeek V4.1 Flash?

Kabuuang parameter, ayon sa [model card ng DeepSeek](https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash): 552B backbone parameters (the card lists 196B of Engram conditional memory separately). Aktibong parameter kada token: 8B per token in prefill, 16B in decode. (sinuri noong Nobyembre 16, 2026; hindi bineripika ng BroadNet ang mga numerong ito.)

### Sinusuportahan ba ng DeepSeek V4.1 Flash ang tool calling?

Yes, per the prompt-encoding reference, ayon sa [model card ng DeepSeek](https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash): test cases for multi-turn conversations, tool calling, thinking mode (sinuri noong Nobyembre 16, 2026). Nakadepende rin sa provider na nagse-serve nito kung gumagana ang tools parameter sa pamamagitan ng API, kaya tingnan mo ang docs ng provider na iyon. Hindi ito sinubukan ng BroadNet.

### Sino ang gumawa ng DeepSeek V4.1 Flash?

Ang DeepSeek V4.1 Flash ay gawa ng DeepSeek. Nasa Hugging Face ang model card nito bilang [`deepseek-ai/DeepSeek-V4.1-Flash`](https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash) (sinuri noong Nobyembre 16, 2026). Mga presyo ng bawat model ng DeepSeek na sinusubaybayan namin: [Presyo ng API ng DeepSeek](https://broadrouter.com/fil/llm-api-pricing/makers/deepseek/).

Subukan ito sa totoong workload mo. May US$5 na panimulang credit ang mga bagong account para sa Gemma at transcription. Kailangan ng top-up para sa DeepSeek, GLM, at Qwen3.8.

[Magsimula](https://inference.broadnet.ai/account/?utm_source=broadrouter&utm_medium=referral&utm_campaign=fil-llm-api-pricing-deepseek-v4.1-flash&utm_content=page-cta) ·

**Mga source**

-   [Alibaba Cloud (sinuri noong Nobyembre 14, 2026)](https://www.alibabacloud.com/help/en/model-studio/model-pricing)
-   [Microsoft Azure (sinuri noong Nobyembre 14, 2026)](https://azure.microsoft.com/en-us/pricing/details/ai-foundry-models/deepseek/)
-   [Baseten (sinuri noong Nobyembre 14, 2026)](https://www.baseten.co/pricing/)
-   [CoreWeave Forge (sinuri noong Nobyembre 15, 2026)](https://coreweave.com/forge-pricing/inference-agents)
-   [DeepInfra (sinuri noong Nobyembre 14, 2026)](https://deepinfra.com/deepseek-ai/DeepSeek-V4.1-Flash)
-   [DigitalOcean (sinuri noong Nobyembre 14, 2026)](https://docs.digitalocean.com/products/inference/details/pricing/)
-   [Doubleword (sinuri noong Nobyembre 16, 2026)](https://docs.doubleword.ai/inference-api/models)
-   [Featherless (sinuri noong Nobyembre 15, 2026)](https://featherless.ai/models/deepseek-ai/DeepSeek-V4.1-Flash)
-   [Fireworks AI (sinuri noong Nobyembre 14, 2026)](https://docs.fireworks.ai/serverless/pricing)
-   [Fireworks AI (sinuri noong Nobyembre 14, 2026)](https://fireworks.ai/pricing)
-   [GMI Cloud (sinuri noong Nobyembre 15, 2026)](https://console.gmicloud.ai/model/llm/deepseek-v4-1-flash/0049392b-5a6b-4e94-acc0-e18b8e293aac)
-   [Google Vertex AI (sinuri noong Nobyembre 14, 2026)](https://cloud.google.com/model-garden/partner-models/pricing)
-   [Hyperbolic (sinuri noong Nobyembre 14, 2026)](https://www.hyperbolic.ai/inference)
-   [Inference.net (sinuri noong Nobyembre 16, 2026)](https://inference.net/models)
-   [Modal (sinuri noong Nobyembre 16, 2026)](https://modal.com/pricing)
-   [Morph (sinuri noong Nobyembre 16, 2026)](https://www.morphllm.com/pricing)
-   [Novita AI (sinuri noong Nobyembre 14, 2026)](https://novita.ai/pricing)
-   [Parasail (sinuri noong Nobyembre 14, 2026)](https://www.parasail.io/pricing)
-   [Reka (sinuri noong Nobyembre 16, 2026)](https://developer.reka.ai/models)
-   [Relace (sinuri noong Nobyembre 16, 2026)](https://relace.ai/pricing)
-   [River (sinuri noong Nobyembre 16, 2026)](https://river.ai/pricing)
-   [SiliconFlow (sinuri noong Nobyembre 14, 2026)](https://www.siliconflow.com/pricing)
-   [Together AI (sinuri noong Nobyembre 14, 2026)](https://www.together.ai/pricing)
-   [Venice (sinuri noong Nobyembre 15, 2026)](https://docs.venice.ai/pricing)
-   [DeepSeek (sinuri noong Nobyembre 14, 2026)](https://api-docs.deepseek.com/quick_start/pricing)
-   [BroadNet Inference, pricing.json (na-sync noong Nobyembre 14, 2026) (Ingles)](https://inference.broadnet.ai/pricing.json?utm_source=broadrouter&utm_medium=referral&utm_campaign=fil-llm-api-pricing-deepseek-v4.1-flash&utm_content=sources)
